Pamukkale Univ Muh Bilim Derg, 25(9), 1056-1061, 2019

(LMSCM’2018-16. Uluslararası Lojistik ve Tedarik Zinciri Kongresi Özel Sayısı)

Pamukkale Üniversitesi Mühendislik Bilimleri Dergisi

Pamukkale University Journal of Engineering Sciences

A model for determining the locations of electric vehicles’ charging stations

in Istanbul
İstanbul’da elektrikli araç şarj istasyonlarının konumlandırılması için bir
model
Büşra Gülnihan DAŞCIOĞLU1* , Gülfem TUZKAYA2 , Hüseyin Selçuk KILIÇ3
1,2,3Department of Industrial Engineering, Engineering Faculty, Marmara University, İstanbul, Turkey.
[email protected], [email protected], [email protected]

Received/Geliş Tarihi: 15.06.2019, Accepted/Kabul Tarihi: 21.11.2019 doi: 10.5505/pajes.2019.28475

* Corresponding author/Yazışılan Yazar Special Issue Article/Özel Sayı Makalesi

Abstract Öz
The studies about Electric Vehicles (EV) have gained importance and Klasik ulaştırma sistemlerinin çevresel kaygıları göz önünde
increased in the last years depending on the environmental concerns of bulundurulduğunda, elektrikli araçlarla (EA) ilgili çalışmalar önem
the classic transportation systems. One of the problems to consider at kazanmış ve son yıllarda sayıca artmıştır. Bu noktada ele alınması
this point is locating the proper points of electric charging stations for gereken sorunlardan biri EA’lar için şarj istasyonlarının uygun
EVs. The primary objective of this research is to locate the electric yerlerinin belirlenmesidir. Bu çalışmanın temel amacı, İstanbul'daki
charging stations in Istanbul considering the flow of the paths. The elektrik şarj istasyonlarına ilişkin en uygun konumları yolların akışını
locations of electric charging stations are determined by using a değerlendirmeye alarak bulmaktır. Şarj istasyonlarının konumları,
mathematical model based on the flow-refuelling location model with yolların kapsanan akışını maksimize etmeyi amaçlayan akış-yakıt
the aim of maximizing the captured flow. The mathematical model is ikmal lokasyon modeline dayanan bir model kullanılarak belirlenmiştir.
run for various values of “p” and the optimum locations are obtained. Matematiksel model farklı “p” değerli için çözülmüş ve en iyi
lokasyonlar belirlenmiştir.
Keywords: Flow-refuelling location model, Assignment problem, Anahtar kelimeler: Akış-yakıt ikmal lokasyon modeli, Atama
Electric vehicles, Electric charging stations problemi, Elektrikli araçlar, Elektrik şarj istasyonları

been examining these special types of vehicles for reducing the

1 Introduction natural damage nevertheless the alternative fuel vehicles and
In recent years, plug-in electric vehicles are receiving attention normal vehicle types are not fully overused [5]. Since the
since they provide good alternatives for liquid fossil fuel current infrastructure is insufficient and the cost of establishing
problems. CO2 emission consumption is calculated around 4% new electric charging stations is high, the advancements in
through the years of 2014 and 2016 [1]. electric vehicles in developing countries such as Turkey occur
slowly. Table 1 shows electric and hybrid automobile sales
According to the survey of Global Carbon Project, Figure 1
numbers [6] in Turkey. For instance, in the first 9 months of
shows CO2 emission from fossil fuels over the last 59 years [2].
2019, 7562 electric and hybrid automobiles are sold in Turkey.
In addition to this gloomy picture, there are also the warnings
of the energy sector about the losses in energy [3]. However, Table 1: Electric and hybrid automobile sales numbers in
researchers prepare studies which aim to decrease carbon Turkey [6].
emissions by increasing the usage of EVs to regulate the source Sales Cumulative Sales
of vehicles using fossil fuel. Year Numbers Numbers
2015 225 225
2016 994 1219
2017 4528 5747
2018 4031 9778
2019 (First 9 months) 7562 17340
Due to increase in the usage level of electric vehicles, demand
on the electric charging stations increases [7]. The primary
challenge of EV charging technology industry is to determine
the optimal capacitated and located electric charging stations
[8]. There are four presumptions made for EVs charging
necessities in Lee and Han’s research (2017) [9]: driver selects
the shortest route, at the beginning of the route, half of the
Figure 1: The amount of CO2 emissions of countries from using battery is full, at the end of the route, the battery must not be
fossil fuels, 1959-2017 [2]. less than half of the battery and finally the travel range is
constant and deterministic.
Type of the battery directly affects the plug-in electric vehicles’
driving performance and cost of the path [4]. Researchers have

1056
 Pamukkale Univ Muh Bilim Derg, 25(9), 1056-1061, 2019
(LMSCM’2018-16. Uluslararası Lojistik ve Tedarik Zinciri Kongresi Özel Sayısı)
B. G. Daşcıoğlu, G. Tuzkaya, H. S. Kılıç

Electric charging station infrastructures are considered in two implemented a hybrid algorithm based on genetic and particle
groups; urban developments which use node-based approach swarm optimization algorithms. After gathering the results, the
[10], this type of electric charging stations meets the demand infrastructure was tested in simulation by real time system.
when the vehicle is in the parking. Intercity developers on the Another multi-criteria method was implemented by
other hand, meet demands of vehicles during long journeys and researchers in [12], for solving the problem in Hungary.
usually use flow-based development approach [11]. In intra- Weighted multi-criteria decision making model was utilized for
city charging demand, the charging process is carried out at the considering the existing stations and installing new stations.
end of the trip of vehicle; in intercity, the process interrupts the The criteria were selected as demographic, economic,
trip of the vehicle [12]. environmental and transportation based.
In this study, a flow-refuelling location model is applied. The In [17], researchers aimed to locate EVCS in three districts of
study is motivated by the fact that the sales rate of alternative Istanbul by using fuzzy AHP methodology. Spatial geographic
vehicles is increasing day-by-day but there are not enough information systems were used for gathering the data and with
stations to cover the demand. To our knowledge, this solution the criteria ranks of fuzzy AHP, a GIS based map layer was
methodology is not applied for Istanbul before. This study created. To solve location problem of EVCS in USA, authors
emphasizes the authorized electric charging station problem extended the set coverage model by adding distance
for plug-in electric vehicles in Istanbul. The goal is to maximize satisfaction function. The main aim of algorithm was to meet
the captured flow of all paths between districts of Anatolian the customer demand by considering the cost constraints and
part of Istanbul. The main contributions of this paper are as battery life of an EV [18]. In a major advance in 2018,
follows: researchers proposed geographic information system based
1. This research extends the previous articles and is the first multi-criteria decision algorithm for selecting the proper
paper that considers all districts in Anatolian part of locations of charging stations in Ankara, Turkey. After
Istanbul, determining criteria and sub-criteria, fuzzy AHP and TOPSIS
2. This study is the first paper that uses flow-refuelling location methodologies were used by considering geographic
problem for the case of Istanbul. information map. The results presented suggested alternative
locations [19].
The first part of the article deals with the fundamentals of the
methods, in the second part, an application is given to In 2019, Ju et al. [20] used multi criteria decision making
investigate the effectiveness of the suggested approach. In the methodology and extended grey relation projection algorithm
last part, the results and conclusions are presented. for ranking the possible locations for EVCS in Beijing, China.
After determining main and sub-criteria, fuzzy AHP and grey
2 Literature review relational projection algorithms were used in order to
Determining the locations of charging station for EVs is the determine the possible locations’ ranks. A data-driven and
most important part of EV studies, therefore most of the studies particle swarm optimization based heuristic algorithm was
consider finding the proper locations of charging stations. developed in order to locate the EVCS with consideration of
There are several different solution methods in this area. round trips [21]. Improved Whale Optimization Algorithm was
modeled by the researchers of [22], for locating EVCS by
The authors in [7], designed electric charging station network considering service risk capacity constraints. The MILP model
in Turkey by using capacitated p-median location model. The and improved Whale Optimization Algorithm were compared
model aimed to maximize the company managers’ preference as conclusion. Both models were aimed to reduce social costs
scores. The capacity data was gathered from managers through without risking the service capacity. A recent paper was
maximum number of districts and maximum possible customer presented by authors of [23], considering the problem of
queue of the location as constraints. In another research, locating fast-charging EVS to the European highway network of
authors [13], proposed three methods for determining France, Germany, the Benelux countries, Switzerland, Austria,
locations; a multi-period optimization model based on flow- Denmark, the Czech Republic, and Poland. Passenger car trips
refueling location model and two myopic methods. They in Europe data were solved by efficient flow-refueling location
considered multi-period plan for Korean Expressways as a case model. The main contribution of the research was considering
study. Performance analysis was made by comparing the the profitability of fast charging stations. Finally, an extended
results of these methods. In another study, bus network of TOPSIS methodology was used for finding the suitable locations
Stockholm was considered for solving the location selection of electric charging stations in Istanbul by Dascioglu et al. [24].
problem of electric charging stations for electric buses. Authors
presented MILP model. The model aimed to minimize 3 Implemented methodology
transportation cost of bus network [14]. The authors in [15], In this section, mathematical representation of flow-refuelling
implemented two-step model by using the data of the flow of location model ([25]-[27]) is explained.
EVs and the road network. First, the model calculated the
required number of electric charging stations and located the The flow capturing problem is originally designed and used for
stations by considering road network of EVs. This method was facility location problem and it is one of the most important
tested in Italian highway network. discrete location theory problems. The flow-refuelling location
model aims to decide the locations of the facilities or stations in
The authors in [16], implemented multi-criteria decision- our situation and allocate the demand points to one or more
making structure for locating charging stations. Criteria were stations by considering the flow of paths’ between locations.
selected by extended sustainability theory and by using fuzzy
Delphi method. In another article, researchers [8], aimed to q: index of OD pairs (the shortest paths for each pair),
solve the optimal siting of electric charging stations problem in Q: set of all OD pairs,
Allahabad, India. For solving the mentioned problem, authors fq: flow volume on the OD pair q,

1057
 Pamukkale Univ Muh Bilim Derg, 25(9), 1056-1061, 2019
(LMSCM’2018-16. Uluslararası Lojistik ve Tedarik Zinciri Kongresi Özel Sayısı)
B. G. Daşcıoğlu, G. Tuzkaya, H. S. Kılıç

k: index of facility locations, The objective of the article is to determine the locations of
K: set of all potential facility locations, electric charging stations. Optimum locations of the districts in
the Anatolian side of Istanbul are obtained by a flow capturing
p: the number of facilities to be located,
method based on the flows between locations and possible
h: index of facility combinations, electric charging station locations.
H: set of all potential facility combinations.
Table 2: Potential station locations.
No Set of All Potential Stations
1, 𝑖𝑓 𝑓𝑞 𝑖𝑠 𝑐𝑎𝑝𝑡𝑢𝑟𝑒𝑑 1 Ataşehir
𝑦𝑞 = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 2 Beykoz
1, 𝑖𝑓 𝑎 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑖𝑠 𝑙𝑜𝑐𝑎𝑡𝑒𝑑 𝑎𝑡 𝑘 3 Çekmeköy
𝑥𝑘 = { 4 Kadıköy
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
1, 𝑖𝑓 𝑎𝑙𝑙 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝑖𝑛 𝑐𝑜𝑚𝑏𝑖𝑛𝑎𝑡𝑖𝑜𝑛 ℎ 𝑎𝑟𝑒 𝑜𝑝𝑒𝑛 5 Kartal
𝑣ℎ = { 6 Maltepe
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 7 Pendik
1, 𝑖𝑓 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑘 𝑖𝑠 𝑖𝑛 𝑐𝑜𝑚𝑏𝑖𝑛𝑎𝑡𝑖𝑜𝑛 ℎ 8 Sancaktepe
𝑎ℎ𝑘 = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 9 Ümraniye
1, 𝑖𝑓 𝑓𝑎𝑐𝑖𝑙𝑖𝑡𝑦 𝑐𝑜𝑚𝑏𝑖𝑛𝑎𝑡𝑖𝑜𝑛 ℎ 𝑐𝑎𝑛 𝑟𝑒𝑓𝑢𝑒𝑙 𝑂𝐷 𝑝𝑎𝑖𝑟 𝑞 10 Üsküdar
𝑏𝑞ℎ = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
Table 3: Combination of facilities.
No Set of All Potential Facility Combinations
(1) 1 Ataşehir-Kadıköy
𝑚𝑎𝑥 𝑍 = ∑ 𝑓𝑞 𝑦𝑞 2 Ataşehir-Maltepe
𝑞∈𝑄 3 Ataşehir - Ümraniye
4 Ataşehir-Üsküdar
(2)
∑ 𝑏𝑞ℎ 𝑣ℎ ≥ 𝑦𝑞 ∀𝑞 ∈ 𝑄 5 Beykoz-Çekmeköy
6 Beykoz-Sancaktepe
ℎ∈𝐻 7 Beykoz-Ümraniye
𝑎ℎ𝑘 𝑥𝑘 ≥ 𝑣ℎ ∀ ℎ ∈ 𝐻; 𝑘 ∈ 𝐾 (3) 8 Beykoz-Üsküdar
9 Çekmeköy-Sancaktepe
(4) 10 Çekmeköy-Şile
∑ 𝑥𝑘 = 𝑝 11 Çekmeköy-Ümraniye
𝑘∈𝐾
12 Kadıköy-Üsküdar
13 Kadıköy-Maltepe
𝑥𝑘 , 𝑣ℎ , 𝑦𝑞 ∈ {0,1} ∀𝑘, ℎ, 𝑞 (5) 14 Kartal-Maltepe
15 Kartal-Pendik
16 Kartal-Sancaktepe
Objective function (1) aims to maximize the flow volume that 17 Kartal-Sultanbeyli
can be refuelled. Constraint (2) prevents the capturing of path 18 Maltepe-Sancaktepe
q, unless a suitable combination of facilities is selected. 19 Maltepe-Sultanbeyli
Constraint (3) ensures the opening of all facilities in a selected 20 Sancaktepe-Sultanbeyli
facility combination h. Constraint (4) ensures that the number 21 Sancaktepe-Tuzla
22 Ataşehir-Ümraniye-Üsküdar
of opened stations equals to p. Constraint (5) provides the 23 Beykoz-Çekmeköy-Sancaktepe
binary constraints. 24 Kadıköy-Maltepe-Kartal
25 Ataşehir-Ümraniye-Üsküdar-Kadıköy
4 Application of the methodology 26 Beykoz-Çekmeköy-Sancaktepe-Sultanbeyli
The suggested approach is applied on nine different policies Paths were created by considering the minimum driving range
considering the number of charging station locations (p) from of a typical electric vehicle. The electric vehicle can travel 40 km
2 to 10. Table 2 shows the potential station locations (k) and with low electric load. In Table 4, the distance matrix of districts
Table 3 shows the combination of facilities (h). While is shown for the Anatolian part of Istanbul.
determining the combination of facilities, it is benefited from
the geographical locations of the districts.
Table 4: Distance matrix of districts.
Sancaktepe

Sultanbeyli
Çekmeköy

Ümraniye

Üsküdar
Ataşehir

Kadıköy

Maltepe
Beykoz

Pendik
Kartal

Tuzla
Şile

District
Ataşehir - 24 34 9 17 11 33 29 14 63 26 7 14
Beykoz 24 - 24 29 37 34 60 35 34 73 47 22 26
Çekmeköy 34 24 - 38 33 27 43 14 21 36 33 29 36
Kadıköy 9 29 38 - 19 13 20 24 18 63 30 12 11
Kartal 17 37 33 19 - 13 7 21 18 60 17 22 27
Maltepe 11 34 27 13 13 - 16 16 11 56 27 19 20
Pendik 33 60 43 20 7 16 - 29 23 71 16 27 29
Sancaktepe 29 35 14 24 21 16 29 - 10 41 21 19 28
Sultanbeyli 14 34 21 18 18 11 23 10 - 50 16 20 25
Şile 63 73 36 63 60 56 71 41 50 - 63 57 63
Tuzla 26 47 33 30 17 27 16 21 16 63 - 32 36
Ümraniye 7 22 29 12 22 19 27 19 20 57 32 - 12
Üsküdar 14 26 36 11 27 20 29 28 25 63 36 12 -

1058
 Pamukkale Univ Muh Bilim Derg, 25(9), 1056-1061, 2019
(LMSCM’2018-16. Uluslararası Lojistik ve Tedarik Zinciri Kongresi Özel Sayısı)
B. G. Daşcıoğlu, G. Tuzkaya, H. S. Kılıç

The data is gathered from General Directorate for Highways The mathematical model is run for the nine different p values
department [28]. Table 5 shows the paths of OD pairs and flows and the optimum solutions are obtained. The results of the
of the paths (fq). The flows are assumed to be proportional with applied model are shown in Table 6. The gap ratio (GR) is
the number of gas stations on the paths. Hence, relative path computed considering the deviation from the maximum
flow scores are obtained based on the path having the minimum captured flow score among the solutions of all p values. For
number of gas stations. instance, the GR for p=2 is obtained via the following
Figure 2 shows a sample path of (number 14) Beykoz - calculations.
Çekmeköy-Sancaktepe-Sultanbeyli on the map. The flow score GR (p = 2) = (358 - 165) / 358 = 53.91%
of this path (15) is obtained by getting the sum of the flow When the GR values are analyzed, it is concluded that there is
scores between its nodes which are given as follows: no significant improvement for the p value bigger than 6. Hence,
Beykoz-Çekmeköy: 6 it does not seem reasonable to establish the system with more
Çekmeköy-Sancaktepe: 5 than 6 electric charging stations in the initial stage.

Sancaktepe-Sultanbeyli: 4 Considering the result of p=2 case, the map of one of the two
selected stations, Kartal, and the captured paths are shown in
Figure 3.

Table 5: Determined OD pairs and the flows of pairs (fq)

No. Path Code Paths Path Flow Score (fq)
1 P1 Ataşehir - Ümraniye 9
2 P2 Beykoz - Çekmeköy 6
3 P3 Çekmeköy - Şile 5
4 P4 Kadıköy - Üsküdar 8
5 P5 Kartal - Pendik 12
6 P6 Maltepe - Sultanbeyli 8
7 P7 Sancaktepe - Tuzla 10
8 P8 Ataşehir - Ümraniye - Üsküdar 13
9 P9 Beykoz - Çekmeköy - Sancaktepe 11
10 P10 Kadıköy - Maltepe - Kartal 12
11 P11 Pendik - Tuzla - Sultanbeyli 17
12 P12 Ataşehir - Ümraniye - Üsküdar - Kadıköy 17
13 P13 Üsküdar - Ümraniye - Çekmeköy - Şile 14
14 P14 Beykoz - Çekmeköy - Sancaktepe - Sultanbeyli 15
15 P15 Maltepe - Kartal - Pendik - Tuzla 21
16 P16 Ataşehir - Ümraniye - Üsküdar - Kadıköy - Maltepe 21
17 P17 Beykoz - Çekmeköy - Sancaktepe - Sultanbeyli - Tuzla 20
18 P18 Ataşehir - Ümraniye - Üsküdar - Kadıköy - Maltepe - Sultanbeyli 25
19 P19 Beykoz - Çekmeköy - Sancaktepe - Sultanbeyli - Pendik -Tuzla 28
20 P20 Beykoz - Üsküdar- Ümraniye - Ataşehir - Kadıköy - Maltepe - Kartal 27
21 P21 Beykoz - Çekmeköy - Sancaktepe - Sultanbeyli - Maltepe - Kartal - Pendik 31
22 P22 Beykoz - Çekmeköy - Sancaktepe - Sultanbeyli - Maltepe - Kartal - Tuzla 28

Figure 2: A sample path between Beykoz-Sultanbeyli. Figure 3: A sample path of selected station Kartal and
covered paths.

1059
 Pamukkale Univ Muh Bilim Derg, 25(9), 1056-1061, 2019
(LMSCM’2018-16. Uluslararası Lojistik ve Tedarik Zinciri Kongresi Özel Sayısı)
B. G. Daşcıoğlu, G. Tuzkaya, H. S. Kılıç

Table 6: Results of the applied model.

p Station Covered Paths Total Flow GR (%)
Kartal P5; P10; P15; P20; P21; P22
2 165 53.91%
Sancaktepe P7; P14; P17; P19; P21; P22
Beykoz P2; P9; P14; P17; P19; P20; P21; P22
3 Çekmeköy P2; P3; P9; P13; P14; P17; P19; P21; P22 258 27.93%
Sancaktepe P7; P9; P14; P17; P19; P21; P22
Beykoz P2; P9; P14; P17; P19; P20; P21; P22
Çekmeköy P2; P3; P9; P13; P14; P17; P19; P21; P22
4 291 18.72%
Kartal P5; P15; P20; P21; P22
Sancaktepe P7; P9; P14; P17; P19; P21; P22
Beykoz P2; P9; P14; P17; P19; P20; P21; P22
Çekmeköy P2; P3; P5; P6; P7; P9; P10; P12; P13; P14; P15; P16; P17; P18; P19; P20; P21; P22
5 Kartal P5; P10; P15; P20; P21; P22 311 13.13%
Maltepe P6; P10; P15; P16; P18; P20; P21; P22
Sancaktepe P7; P9; P14; P17; P19; P21; P22
Ataşehir P1; P8; P12; P16; P18; P20
Çekmeköy P3; P9; P13; P14; P17; P19; P21; P22
Kartal P5; P10; P15; P20; P21; P22
6 327 8.66%
Maltepe P6; P10; P15; P16; P18; P20; P21; P22
Sancaktepe P7; P9; P14; P17; P19; P21; P22
Ümraniye P1; P8; P12; P13; P16; P18; P20
Ataşehir P1; P8; P12; P16; P17; P18; P20
Beykoz P2; P9; P14; P17; P19; P20; P21; P22
Çekmeköy P2; P3; P9; P13; P14; P17; P19; P21; P22
7 Kartal P5; P10; P15; P20; P21; P22 333 6.98%
Maltepe P6; P10; P15; P16; P18; P20; P21; P22
Sancaktepe P7; P9; P14; P17; P19; P21; P22
Ümraniye P1; P8; P12; P13; P16; P18; P20
Ataşehir P1; P8; P12; P16; P18; P20
Çekmeköy P3; P9; P13; P14; P15; P17; P19; P21; P22
Kadıköy P4; P10; P12; P16; P18; P20
Kartal P5; P10; P15; P20; P21; P22
8 352 1.68%
Maltepe P6; P10; P15; P16; P18; P20; P21; P22
Sancaktepe P7; P9; P14; P17; P19; P21; P22
Ümraniye P1; P8; P12; P13; P16; P18; P20
Üsküdar P4; P8; P12; P13; P16; P18; P20
Ataşehir P1; P8; P12; P16; P18; P20
Beykoz P2; P9; P14; P17; P19; P20; P21; P22
Çekmeköy P2; P3; P9; P13; P14; P17; P19; P21; P22
Kadıköy P4; P10; P12; P16; P18; P20
9 Kartal P5; P10; P15; P20; P21; P22 358 0%
Maltepe P6; P10; P15; P16; P18; P20; P21; P22
Sancaktepe P7; P9; P14; P17; P19; P21; P22
Ümraniye P1; P8; P12; P13; P16; P18; P20
Üsküdar P4; P8; P12; P13; P16; P18; P20
Ataşehir P1; P8; P12; P16; P18; P20
Beykoz P2; P9; P14; P17; P19; P20; P21; P22
Çekmeköy P2; P3; P9; P13; P14; P17; P19; P21; P22
Kadıköy P4; P10; P12; P16; P18; P20
Kartal P5; P10; P15; P20; P21; P22
10 358 0%
Maltepe P6; P10; P15; P16; P18; P20; P21; P22
Pendik P5; P11; P15; P19; P21
Sancaktepe P7; P9; P14; P17; P19; P21; P22
Ümraniye P1; P8; P12; P13; P16; P18; P20
Üsküdar P4; P8; P12; P13; P16; P18; P20

5 Conclusion handled in this study. Flow-refuelling location model is used as

the solution methodology and an application is performed in
One of the main concerns of academic and practical world is the Anatolian part of Istanbul so as to maximize the total
about reducing the level of global warming. Due to the captured flow rate. The stations and paths captured by stations
importance of the topic, various applications are being are obtained for different p values.
performed in an increasing trend. However, electric vehicles
play a significant and direct role for overcoming the global The location problem of electric charging stations is handled in
warming and are the focus of this study. Specifically, the a classic way in this study. However, for further studies,
problem of determining electric charging stations’ locations is additional constraints can be added by considering the capacity

1060
 Pamukkale Univ Muh Bilim Derg, 25(9), 1056-1061, 2019
(LMSCM’2018-16. Uluslararası Lojistik ve Tedarik Zinciri Kongresi Özel Sayısı)
B. G. Daşcıoğlu, G. Tuzkaya, H. S. Kılıç

and type of stations. Moreover, for large-size problems, [16] Zhao H, Li N. “Optimal siting of charging stations for
heuristic algorithms can be developed. Last but not least, multi- electric vehicles based on fuzzy delphi and hybrid multi-
objective approaches can be applied considering various criteria decision making approaches from an extended
objectives simultaneously. sustainability perspective”. Energies, 9(4), 1-22, 2016.
[17] Guler D, Yomralioglu T. “GIS and fuzzy AHP based area
6 References selection for electric vehicle charging stations”. ISPRS TC
[1] Joint Research Centre. “Emissions Data and Maps”. IV Mid-term Symposium on 3D Spatial Information Science
https://edgar.jrc.ec.europa.eu (01.07.2018). – The Engine of Change, Delft, The Netherlands, 01-05
[2] Olivier, JGJ, Schure KM, Peters JAHW. “Trends in global October 2018.
CO2 and total greenhouse gas emissions”. PBL [18] Afroditi M, Boile S, Theofanis ES, Margaritis D. “Electric
Netherlands Environmental Assessment Agency Report, Vehicle Routing Problem with industry constraints: trends
2017. and insights for future research”. Transportation Research
[3] Nakata T. “Analysis of the impact of plug-in hybrid Procedia, 3, 452-459, 2014.
vehicles on energy systems in Japan”. Transportation [19] Erbas M, Kabak M, Ozceylan E. “Optimal siting of electric
Research: Part D, 5, 373-383, 2017. vehicle charging stations: A GIS-based fuzzy Multi-Criteria
[4] Rahman I, Vasant PM, Singh BSM, Abdullah-Al-Wadud M, Decision Analysis”. Energy, 163, 1017-1031, 2018.
Adnan N. “Review of recent trends in optimization [20] Ju Y, Ju D, Gonzalez EDRS, Giannakis M, Wang A. “Study of
techniques for plug-in hybrid, and electric vehicle site selection of electric vehicle charging station based on
charging infrastructures”. Renewable and Sustainable extended GRP method under picture fuzzy environment”.
Energy Reviews, 58, 1039-1047, 2016. Computer&Industrial Engineering, 135, 1271-1285, 2019.
[5] Iwata K, Matsumoto S. “Use of plug-in hybrid vehicles in [21] Liu Q, Liu J, Le W, Guo Z, He Z. “Data-driven intelligent
japan: an analysis of used car market data”. Transportation location of public charging stations for electric vehicles”.
Research, Part D, 46, 200-206, 2016. Journal of Cleaner Production, 232, 531-541, 2019.
[6] Türkiye Elektrikli ve Hibrit Araçlar Derneği. “Makaleler”. [22] Zhang Y, Qiu Z, Gao P, Jiang S. “Location model of electric
http://tehad.org (01.10.2018). vehicle charging stations”. International Conference on
[7] Gavranovic H, Barut A, Ertek G, Yüzbaşıoğlu OB, Physics, Mathematics and Statistics, Shanghai, China, 12-14
Pekpostalcı O, Tombuş Ö. “Optimizing the electric charge May 2018.
station network of EŞARJ”. Procedia Computer Science, 31, [23] Jochem P, Szimba E, Oppermann M. “How many fast-
15-21, 2014. charging stations do we need along European highways?”.
[8] Awasthi A, Venkitusamy K, Padmanaban P, Transportation Research, Part D, 73, 120-129, 2019.
Selvamuthukumaran R, Blaabjerg F, Singh AK. “Optimal [24] Dascıoglu BG, Kalender ZT, Tuzkaya G, Kilic HS. Evaluation
planning of electric vehicle charging station at the of Electric Vehicles Station Locations with an Extended
distribution system using hybrid optimization algorithm”. TOPSIS Methodology Using Probabilistic Linguistic Term
Energy, 133, 70-78, 2017. Sets. Editors: Kahraman C, Cebi S, Cevik Onar S, Oztaysi B,
[9] Lee C, Han J. “Benders-and-Price approach for electric Tolga A, Sari I. Intelligent and Fuzzy Techniques in Big
vehicle charging station location problem under Data Analytics and Decision Making, 820-828, Springer,
probabilistic travel range”. Transportation Research, Cham, 2019.
Part B, 106, 130-152, 2017. [25] Kuby MJ, Lim S. “The flow-refueling location problem for
[10] Hakimi SL. “Optimum locations of switching centers and alternative-fuel vehicles”. Socio-Economic Planning
the absolute centers and medians of a graph”. Operations Science, 39, 125-145, 2005.
Research, 12(3), 450-459, 1964. [26] Kuby MJ, Lim S. “Location of alternative-fuel stations using
[11] Hodgson MJ. “A flow-capturing location-allocation model”. the Flow-Refueling Location Model and dispersion of
Geographical Analysis, 22(3), 270-279, 1990. candidate sites on arcs”. Network and Spatial Economics,
[12] Csonka B, Csiszá r C. “Determination of charging 7(2), 129-152, 2007.
infrastructure location for electric vehicles”. [27] Kuby MJ, Lines L, Schultz R, Xie Z, Kim JG, Lim S.
Transportation Research Procedia, 27, 768-775, 2017. “Optimization of hydrogen stations in Florida using the
[13] Chung SH, Kwon C. “Multi-period planning for electric car flow-refueling location model”. International Journal of
charging station locations: A case of Korean Expressways”. Hydrogen Energy, 34, 6045-6064, 2009.
European Journal of Operational Research, 242, 677-687, [28] General Directorate for Highways department.
2015. “İstatistikler”. https://www.kgm.gov.tr/.
[14] Xylia M, Leduc S, Patrizio P, Kraxner F, Silveria S. “Locating
charging infrastructure for electric buses in Stockholm”.
Transportation Research, Part C, 78, 183-200, 2017.
[15] Polimeni A, Napoli G, Andaloro L, Antonucci V. “Electric
vehicle charging infrastructure planning in a road
network”. Renewable and Sustainable Energy Reviews,
80, 98-108, 2017.

1061